An extension to the brightness clustering transform and locally contrasting keypoints
An extension to the brightness clustering transform and locally contrasting keypoints
The need for faster feature matching has left as a result a new set of feature descriptors to the computer vision community, ORB, BRISK and FREAK amongst others. These new descriptors allow reduced time and memory consumption on the processing and storage stages, mitigating the implementation of more complex tasks. The problem is now the lack of fast interest point detectors with good repeatability to use with these new descriptors. A blob-detection algorithm was recently presented that uses an innovative non-deterministic low-level operator called the Brightness Clustering Transform (BCT) (Lomeli-R. and Nixon in The brightness clustering transform and locally contrasting keypoints. In CAIP. Springer, Berlin, pp 362–373, 2015). This algorithm is easy to implement and is faster than most of the currently used feature detectors. The BCT can be thought as a coarse-to-fine search through scale spaces for the true derivative of the image. The new algorithm is called Locally Contrasting Keypoints detector (LOCKY). Showing good robustness to image transformations included in the Oxford affine-covariant regions dataset, LOCKY is amongst the fastest affine-covariant feature detectors. In this paper, we present an extension of the BCT that detects larger structures maintaining timing and repeatability; this extension is called the BCT-S.
1187–1196
Lomeli-R., Jaime
26137049-8361-4a68-a7d9-734befbc5e58
Nixon, Mark
2b5b9804-5a81-462a-82e6-92ee5fa74e12
November 2016
Lomeli-R., Jaime
26137049-8361-4a68-a7d9-734befbc5e58
Nixon, Mark
2b5b9804-5a81-462a-82e6-92ee5fa74e12
Lomeli-R., Jaime and Nixon, Mark
(2016)
An extension to the brightness clustering transform and locally contrasting keypoints.
[in special issue: CAIP 2015]
Machine Vision and Applications, 27, .
(doi:10.1007/s00138-016-0785-3).
Abstract
The need for faster feature matching has left as a result a new set of feature descriptors to the computer vision community, ORB, BRISK and FREAK amongst others. These new descriptors allow reduced time and memory consumption on the processing and storage stages, mitigating the implementation of more complex tasks. The problem is now the lack of fast interest point detectors with good repeatability to use with these new descriptors. A blob-detection algorithm was recently presented that uses an innovative non-deterministic low-level operator called the Brightness Clustering Transform (BCT) (Lomeli-R. and Nixon in The brightness clustering transform and locally contrasting keypoints. In CAIP. Springer, Berlin, pp 362–373, 2015). This algorithm is easy to implement and is faster than most of the currently used feature detectors. The BCT can be thought as a coarse-to-fine search through scale spaces for the true derivative of the image. The new algorithm is called Locally Contrasting Keypoints detector (LOCKY). Showing good robustness to image transformations included in the Oxford affine-covariant regions dataset, LOCKY is amongst the fastest affine-covariant feature detectors. In this paper, we present an extension of the BCT that detects larger structures maintaining timing and repeatability; this extension is called the BCT-S.
Text
An Extension to the Brightness Clustering Transform and Locally Contrasting Keypoints Rev-2.pdf
- Accepted Manuscript
Available under License Other.
Text
s00138-016-0785-3
- Version of Record
More information
Accepted/In Press date: 19 May 2016
e-pub ahead of print date: 9 July 2016
Published date: November 2016
Organisations:
Vision, Learning and Control
Identifiers
Local EPrints ID: 396009
URI: http://eprints.soton.ac.uk/id/eprint/396009
ISSN: 0932-8092
PURE UUID: 5aa9f7b1-1c9c-4956-a157-a2812e16ac33
Catalogue record
Date deposited: 02 Jun 2016 10:48
Last modified: 15 Mar 2024 05:37
Export record
Altmetrics
Contributors
Author:
Jaime Lomeli-R.
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
